Wave Boundary Layer Model based wind drag estimation for tropical storm surge modelling in the Bay of Bengal

Abstract

Modelling of storm surge numerically has always been strenuous as it is associated with various influencing factors. The major governing component being the wind forcing or the wind stress - that signifies the computational accuracy of simulated surge and wave parameters. The present study is aimed at analysing the most reliable wind drag evaluation method for real-time predictions of storm surge. In an attempt to examine the same, two tropical cyclones (Thane and Vardah) eventuated in the Northern Indian Ocean were modelled. Four most widely used linear empirical and quadratic relations along with the enhanced Wave Boundary Layer Model (WBLM) were executed for the estimations of wind drag coefficient. The surge was subsequently simulated (using the coupled hydrodynamic circulation and wave model: ADCIRC and SWAN, respectively), individually for each of the above wind stress methods to obtain the corresponding water levels and wave features. The modelled values were further validated with the observed water level and wave data. It was quite intuitively observed that, WBLM outperformed (i.e. significantly correlated with the observed values) its linear counterparts since, the former pragmatically includes the effects of air-sea interface at high wind speeds in the model. However, due to insufficiency of the observed parameters, water levels obtained at locations of maximum wind speeds could not be satisfactorily validated. The WBLM-based computation of significant wave heights in deep as well as shallow water nevertheless enabled efficient and reasonably-reliable estimations of the peak incidents (along with relative wave directions and energy densities). Although this study essentially includes cyclonic wind velocities, it was identified that - with given wind field that includes the ‘pre-existing’ wind-wave climate in the relevant domain (i.e. before the cyclone genesis) the modelled estimates could be more precise.